Filopodia are actin-based finger-like protrusions from the plasma membrane, and are used for multiple functions in cells including motility, cell-cell adhesion, cell-substratum adhesion, and viral infection. For many of these processes, the filopodia involved are highly dynamic, assembling and disassembling on a time scale of minutes. At least three actin-based activities are required for filopodial assembly: filament nucleation activity, filament elongation activity, and filament bundling activity. Two competing models for filopodial assembly differ in the nature of the nucleation and elongation activities. In the tip nucleation model, proteins such as formins act as both nucleation and elongation factors. In the convergent elongation model, Arp2/3 complex is the nucleation factor, with formins subsequently acting as elongation factors that also help re-model the branched Arp2/3-generated actin network. Our results show that the formin FMNL3 is a potent filopodial assembly factor, and we propose a novel extension to the convergent elongation model - that FMNL3 can remodel any existing filaments (Arp2/3-dependent or Arp2/3-independent) to filopodia, provided they abut the plasma membrane. In addition, we show that FMNL3 acts in cell-cell adhesion. In this proposal, we study FMNL3 in mammalian cells, focusing on the following aims. Aim 1 uses cell-based assays to define the mechanism of FMNL3-mediated filopodial assembly. We use live-cell microscopy, inhibitor treatments and siRNA to test the ability of FMNL3 to re-model stress fiber/focal adhesion-associated actin filaments into filopodia. Aim 2 uses a cell-free system to reconstitute filopodial assembly on supported lipid bilayers using purified proteins (FMNL3, Arp2/3 complex, capping protein, profilin, fascin). With this system, we will test assembly principles in a controlled manner and investigate the contributions of other molecules (VASP, cofilin). Aim 3 investigates FMNL3 function in cell-cell adhesion, focusing on FMNL3's transit from intracellular storage sites to the plasma membrane, and activation at the plasma membrane during this process. Overall, this project will provide fundamentally novel mechanistic information on filopodial assembly, as well as providing novel molecular connections between actin dynamics and early events in cell-cell adhesion.